Tetrasodium hydrogen 2-phosphonatobutane-1,2,4-tricarboxylate

Phototransformation in water

Link to relevant study record(s)

Reference

Endpoint:

phototransformation in water

Type of information:

experimental study

Adequacy of study:

key study

Study period:

11/93 - 08/96

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

guideline study with acceptable restrictions

Study type:

direct photolysis

Qualifier:

according to guideline

Guideline:

other: Phototransformation of Chemicals in Water, Part A: Direct Phototransformation. UBA, Berlin, FRG (December 1992).

Deviations:

no

GLP compliance:

yes

Radiolabelling:

yes

Analytical method:

high-performance liquid chromatography

Buffers:

0.02 M acetate buffer pH 5,
0.02 M phosphate buffer pH 7,
0.02 M borate buffer pH 9

Light source:

Xenon lamp

Light spectrum: wavelength in nm:

>= 295 - <= 490

Details on light source:

Merry-go-round irradiation apparatus 13/150, Mangels Co., with Hg-lamp TQ 150, Original Hanau Co., and with Duran 50 filter finger (for spectral data of the lamp used in case of the filtration through Duran 50 glass see Appendix 4).
Thermostatization by circulation condenser 7007 (Huber Co.)

Details on test conditions:

The degradation experiment was conducted in a merry-go-round irradiation apparatus. This is fitted with a mercury immersion lamp TQ 150 of Original Hanau Co.. When using this artificial light source, the higher-energy UV-rays (wave length < 295 nm) are almost quantitatively absorbed by incorporation of a Duran 50 filter tube.
The intensity of the light acting on the test solution was measured by means of the chemical actinometer uranyl oxalate in the cuvettes which were also used in the respective degradation experiment. The uranyl oxalate actinometer is particularly suitable for the high light intensity occurring in the test set-up on account of its relatively low sensitivity in the spectral range from 300 to 490 nm. Relatively complex arithmetic operations are necessary in order to be able to calculate the total amount of radiation entering the measuring cuvette from the number of photons being absorbed by the actinometer. These computations are carried out by means of the computer program Quant.

Duration:

20 min

Temp.:

25 °C

Initial conc. measured:

0.98 mg/L

Duration:

420 min

Temp.:

25 °C

Initial conc. measured:

1 mg/L

Reference substance:

yes

Remarks:

2-Phosphono[3, 4-14C]butane-1,2,4-tricarboxylic acid

Computational methods:

see Attachment "37971-36-1_Photolysis_calculation.pdf", Computational Methods

Preliminary study:

No data provided.

Test performance:

UV-Vis absorption spectrum of 20.5 mg PBTC-Na2 / L pure water stored in brown glass vessel for 24h.
(UV-Vis absorption spectrum of 20.5 mg PBTC-Na2 / L pure water stored in white glass vessel showed not any absorption of light above 240 nm).

Parameter:

epsilon 295 nm

Value:

46 L/(mol cm)

Remarks:

UV-Vis absorption spectrum of 20.5 mg PBTC-Na2 / L pure water stored in brown glass vessel for 24h. (UV-Vis absorption spectrum of 20.5 mg PBTC-Na2 / L pure water stored in white glass vessel showed not any absorption of light above 240 nm).

Quantum yield (for direct photolysis):

1.84

Transformation products:

yes

Details on results:

UV-VIS absorption properties:

The UV-Vis absorption spectrum of 20.5 mg PBTC-Na2 / L pure water stored in white glass vessel showed not any absorption of light above 240 nm (see for Appendix 8). However, the UV-Vis absorption spectrum of 20.5 mg PBTC-Na2 / L pure water stored in brown glass vessel for 24 hours showed a very low but measurable absorption of light (295 nm = 46 L/mole cm) extending into the environmentally relevant wavelengths to about 327 nm. The UV-Vis absorption spectra of 20.5 mg PBTC-Na2  / L in buffers pH 5, 7 and 9 stored in white glass vessels showed not any absorption of light above 250 nm. The UV-Vis absorption spectrum of 20.5 mg PBTC-Na2  + 20.83 mg FeCl3 x 6 H2O / L pure water showed an absorption maximum at 215 nm and a shoulder at about 274 nm. The absorption of light (295 nm = 766 L/mole cm) extended into the environmentally relevant wavelengths to about 322 nm. However, a negative absorption (i.e. an emission of light quanta) was measurable above 322 nm with an emission maximum at about 370 nm (band width from about 340 to 405 nm). The UV-Vis absorption spectrum of 20.5 mg PBTC-Na2  + 20.83 mg FeCl3 x 6 H2O / L 0.02 M buffer pH 5 showed an absorption maximum at 212 nm and a shoulder at about 271 nm. The absorption of light extended into the environmentally relevant wavelengths to about 328 nm. However, a negative absorption (i.e. an emission of light quanta) was measurable above 328 nm with an emission maximum at about 370 nm (band width from about 345 to 410 nm).

Direct interactions of PBTC in water with the sunlight in the troposphere are only possible if e. g. Fe(III) or other ions (i.e. that being liberated from brown glass) are present. The greatest absorption cross-section was measured for PBTC-Na2  + FeCl3 x 6 H2O in buffer pH 9.

Photodegradation in the merry-go-round irradiation apparatus
Intensity of irradiation

The actinometer determination showed a titration difference of 0.70 mL (exp. #1), 0.69 mL (expts. #2 and #3), 0.63 mL (exp. #4) and 0.65 mL (exp. #5) for the respective degradation experiments. In case of experiments #3 and #4, where the quantum yields were calculated, an intensity of the radiation absorbed by the actinometer of 6.30 x 10E+16 and 5.75 x 10E+16 photons per sec. and 3 mL resulted in the range of 295 to 490 nm.

Results with reference substance:

The corresponding irradiation of [14C]PBTC in a natural water (Hönniger Weiher, NRW) resulted in a PBTC half-life of about 30 minutes and a PBTC steady state concentration of less than 10% of the initial concentration of 1 mg/L. This proved that photolysis can contribute to the overall elimination of PBTC in natural waters. The measured fast photo-transformation was unexpected, because not any absorption of light above 230 nm was measurable. Therefore, not any calculation of a quantum yield according to the above-mentioned method was possible. Nevertheless, the acting quantum yield must have been extremely high (yield >> 1), because that fast photo-transformation was measured.

Kinetics of degradation

The experiments in buffer pH 5 (exp. #1 and #2) showed a fast decline of PBTC to about 50% of initial concentration within 4 minutes. But the analysis of the longer irradiated samples (exp. #1: 20 min.; exp. #2: 60 min.) indicated that a kind of steady state photosystem (equilibrium conc.) in the range of PBTC: product = 55:45 to about 35:65 resulted. Therefore, it was neither possible to define a rate constant (or degradation curve), nor to calculate the quantum yield of direct photodegradation in that solution.

The analytical results of experiment #1 were listed in the following table.

Experiment #1
Duration of irradiation [min.]	Concentration of PBTC [mg/L]
0	0.98
2	0.61
4	0.50
6	0.50
8	0.52
10	0.54
12	0.54
14	0.53
16	0.56
18	0.55
20	0.47

The experiment in buffer pH 9 (exp. #3) showed a less fast but more or less constant decrease of the PBTC peak to 28% of initial concentration within 420 minutes. On basis of this degradation a quantum yield was calculated (see below: Identification of degradation products).

The respective data are summarized in the left part of the following table.

Experiment #3		Experiment #4
Duration of	Concentration of	Duration of	Concentration of
irradiation	PBTC	irradiation	PBTC
[ min. ]	[ mg/L ]	[ min. ]	[ mg/L ]
0	1.00	0	1.12
12	0.95	3	0.96
24	0.97	6	0.80
30	0.81	9	0.82
42	0.80	12	0.47
48	0.79	15	0.33
54	0.79	18	0.23
60	0.74	21	0.17
300	0.38	24	0.11
420	0.28	(27)*	(0.34)*
-	-	(30)*	(0.16)*
Statistical evaluation	Experiment #3	-	Experiment #4
No. of data pairs	10	-	9
Rate constant (k)	0.003 min-1	-	0.0993 min-1
Half-life (DT 50)	234.1 min	-	7.0 min
t10% (DT 10)	35.6 min	-	1.1 min
Determination coeff.	0.9828	-	0.9595
Correlation coeff.(R)	-0.9914	-	-0.9796

*: not included for evaluations

Experiment #4 (PBTC in pure water stored in brown glass) showed a much faster decrease of PBTC peak to about 10% of initial concentration within 24 minutes. As it was observed in the experiments #1 and #2 the analytical results of the samples exposed for 27 and 30 min. indicated that the irradiation did not lead to a zero concentration of PBTC. On basis of the degradation curve from 00-24 minutes a quantum yield was calculated. The respective data were summarized in the right part of the previous table.

Experiment #5 (PBTC in natural water) showed a fast transformation of PBTC to the main photoproduct with a half-life of about 32 minutes. In this case and after an irradiation for 420 minutes, either no equilibrium between PBTC and photoproduct occurred or the remaining portion of PBTC was below the quantification limit (about 0.10 mg PBTC / L).

The measured fast degradation was unexpected, because not any absorption of light with wavelengths higher than 230 nm was measurable, even in the high concentrated solution containing 20.45 mg PBTC /L prepared for recording of the UV-VIS spectrum. Therefore, not any calculation of a quantum yield according to the ECETOC method was possible. Nevertheless, the acting quantum yield must have been extremely high (yield >> 1) when considering the quotient for yield ( yield = no. of reacted (degraded) molecules per unit of time / total no. of absorbed light quanta per unit of time), because the total number of absorbed light quanta approached zero. The respective data were summarized in the following table.

Experiment #5
Duration of irradiation [min.]	Concentration of PBTC [mg/L]
0	1.13
20	0.72
30	0.52
45	0.42
60	0.23
90	0.17
135	0.10
180	0.10
225	0.10
255	0.10
360	0.10
390	0.10
420	0.10

6.2.3                Identification of degradation products

Great efforts in isolation and identification of the main photoproduct of PBTC by means of MS and ¹H-NMR were not successful. In principle, the separation of the main photoproduct (Rt  = 3.95 - 4.15 min.) from PBTC peak (Rt  = 3.20 - 3.30 min.) by taking the respective HPLC fractions was possible. But the fractions of main metabolite could not be handled, purified and investigated successfully. Only traces of butane-1,2,4-tricarboxylic acid (BTC) could be found by means of GC-MS investigation of the methylated product fraction. In HPLC the zone of BTC reference was eluted quite in the same range as the main photoproduct with a peak at about Rt  = 4.15 - 4.25 min. The poor chromatographic separation was responsible for the lack to calculate the degradation kinetics as well as the quantum yield of direct photodegradation of PBTC versus the formation of BTC, which was found to be a final photo-product of the photolysis of PBTC.

Exp. #3:

Calculation of the Quantum yield using polychromatic light

Degradation experiment in the merry-go-round apparatus

Exposed volume	3.0 (mL)
Intensity of the radiation absorbed by the actinometer in the range 295 to 490 nm	1.27870E+19 (photons/sec)
Exposure period of the actinometric determination	10.0 (min)
Titration difference for 3.0 mL of the exposed and unexposed actinometer solution	0.689999 (mL)
Intensity of the radiation being absorbed by the actinometer in the range 295 to 490 nm	6.29572E+16 (photons/sec/3mL)
Preparation of the solution for the degradation experiment: Weighed amount Volume of the solution Solvent	1.00 (mg) 1,000.00 (mL) water
Concentration of the test substance in the degradation experiment : c1 c(10%)	1.0  mg/L 6.69111E+14 molecules/ 3 mL
Time for the degradation of 10% of the test substance	35.6 (min)

UV-Spectrum of the test substance

Solvent	0.02 M buffer pH 9
Concentration	17.6 mg/L (Na content was abstracted)

Quantum yield of the photodegradation : 2.17309E-03

Exp. #4: Calculation of the Quantum yield using polychromatic light - Degradation experiment in the merry-go-round apparatus

Exposed volume	3.0 (mL)
Intensity of the radiation absorbed by the actinometer in the range 295 to 490 nm	1.27870E+19 (photons/sec)
Exposure period of the actinometric determination	10.0 (min)
Titration difference for 3.0 mL of the exposed and unexposed actinometer solution	0.629999 (mL)
Intensity of the radiation being absorbed by the actinometer in the range 295 to 490 nm	5.74827E+16 (photons/sec/ 3 mL)
Preparation of the solution for the degradation experiment: Weighed amount Volume of the solution Solvent	1.12 (mg) 1,000.00 (mL) water
Concentration of the test substance in the degradation experiment : c1 c(10%)	1.12 mg/L 7.49404E+14 molecules/ 3 mL
Time for the degradation of 10% of the test substance	1.1 (min)

UV-Spectrum of the test substance

Solvent	water (brown glass)
Concentration	17.6 mg/L  (Na content was abstracted)

Quantum yield of the photodegradation : 1.8408

Validity criteria fulfilled:

not specified

Conclusions:

The irradiation of [14C]PBTC (2-Phosphono[3, 4-14C]butane-1,2,4-tricarboxylic acid) with simulated sunlight (artificial light source) in a merry-go-round apparatus showed different results. Generally, the formation of a reactive photosystem, probably by complexation of ions [e.g. Fe (III)], was found to be necessary for an absorption of environmentally relevant light (wave length > 290 nm) as well as for a transformation of PBTC to a slightly less polar main photoproduct.
Depending on the marginal conditions in the irradiated solutions [pH and ion concentration, e.g. Fe (III)] an equilibrium between PBTC and photoproduct at quite different ratios was to be observed. The most effective transformation at a low level of remaining PBTC in the irradiated solutions was determined for pH 9 in the presence of Fe(III), but still more transformation was observed in pure water stored together with PBTC in a brown glass vessel prior to irradiation. Higher amounts of ions being present in the test solution, e. g. Fe (III), probably decreased the transformation rate or enhanced the back-reaction to PBTC. This influence was already described by Kleinstück [Degradation of phosphonic acids in aqueous solution by light (KSK 4410). BAYER AG, Leverkusen, AC, unpublished report no. AC-6-1252-/m (1990).]. The butane-1,2,4-tricarboxylic acid (BTC) was found to be a final product of PBTC photolysis, but the amounts at the respective sampling periods could not be measured, because a poor chromatographic separation from the main photoproduct. Therefore, no further information about the stability of the main photoproduct could be given.

The corresponding irradiation in natural water (Hönniger Weiher) resulted in a PBTC half-life of about 30 minutes and a PBTC steady state concentration of less than 10% of initial after 420 minutes. This proved that photolysis can contribute to the overall degradation of PBTC in natural waters.

The estimates of "environmental photolysis half-lives" based on two different arithmetic models (GC-SOLAR and Frank & Klöpffer) by means of the resulting quantum yields and the light absorption data in the environmentally relevant range of wavelengths were well comparable when considering identical marginal conditions. The results of modelling based on the irradiation of PBTC in buffer pH 9 in presence of FeCl3 indicated that the mean photolysis half-life should range from 2-3 days in summer to 15-65 days in winter. The results of modelling based on the irradiation of PBTC in pure water, stored in brown glass prior to irradiation indicated that the mean photolysis half-life should range from 0.2-0.3 days in summer to 1-10 days in winter.
The before-mentioned assessments did not consider any indirect photodegradation mechanisms which could increase the degradation in natural waters.

Executive summary:

For a computer-based assessment of the so-called environmental photolysis half-life in water the quantum yield of direct photoreaction of [3,4-14C]PBTC should be determined according to the ECETOC method in polychromatic light.

The irradiation of [14C]PBTC with simulated sunlight (filtered Xenon light) in a merry-go-round apparatus showed different results. Generally, the formation of a reactive photosystem, probably by complexation of ions [e.g. Fe(III)], was found to be necessary for an absorption of environmentally relevant light (yield > 290 nm) as well as for a transformation of PBTC to a slightly less polar main photoproduct.

Depending on the marginal conditions of irradiation [pH and ion concentration, e.g. Fe(III)] an equilibrium between PBTC and main photoproduct at quite different ratios was observed. An effective transformation at a low level of remaining PBTC in the irradiated solutions was determined for pH 9 in the presence of Fe(III). The quantum yield was calculated to be 0.0022. More effective was the transformation in pure water being stored together with PBTC in a brown glass vessel prior to irradiation. In that case, only traces of e.g. Fe or Mn ions can have been dissolved, but the quantum yield was calculated to be 1.84. Higher amounts of ions being present in the test solution, e.g. Fe(III), probably decreased the transformation rate or enhanced a back-reaction to PBTC.

The butane-1,2,4-tricarboxylic acid (BTC) was found to be a final product of photolysis of PBTC. The quantities of BTC at the respective sampling periods could not be determined, because of a bad chromatographic separation from the main photoproduct. Therefore, no further information about stability of the main photoproduct could be given.

The corresponding irradiation of [14C]PBTC in a natural water (Hönniger Weiher, NRW) resulted in a PBTC half-life of about 30 minutes and a PBTC steady state concentration of less than 10% of the initial concentration of 1 mg/L. This proved that photolysis can contribute to the overall elimination of PBTC in natural waters. The measured fast photo-transformation was unexpected, because not any absorption of light above 230 nm was measurable. Therefore, not any calculation of a quantum yield according to the above-mentioned method was possible. Nevertheless, the acting quantum yield must have been extremely high (yield >> 1), because that fast photo-transformation was measured.

The estimates of "environmental photolysis half-lives" based on two different arithmetic models (GC-SOLAR and Frank & Klöpffer) by means of the resulting quantum yields and the light absorption data in the environmentally relevant range of wavelengths were well comparable when considering identical marginal conditions. The results of modelling based on the irradiation of PBTC in buffer pH 9 and in presence of FeCl3 indicated that the mean photolysis half-life should range from 2-3 days in summer to 15-65 days in winter. The results of modelling based on the irradiation of PBTC in pure water and stored in brown glass prior to irradiation indicated that the mean photolysis half-lives should range from 0.2-0.3 days in summer to 1-10 days in winter.

Description of key information

As there is no data available for tetrasodium hydrogen 2-phosphonatobutane-1,2,4-tricarboxylate, results of the parent acid, 2-phosphonobutane-1,2,4-tricarboxylic acid are taken into account for this endpoint.
The results of modelling based on the irradiation of 2-phosphonobutane-1,2,4-tricarboxylic acid in buffer pH 9 and in presence of FeCl3 indicated that the mean photolysis half-life should range from 2-3 days in summer to 15-65 days in winter. The results of modelling based on the irradiation of 2-phosphonobutane-1,2,4-tricarboxylic acid in pure water and stored in brown glass prior to irradiation indicated that the mean photolysis half-lives should range from 0.2-0.3 days in summer to 1-10 days in winter.

Key value for chemical safety assessment

Half-life in water:

10 d

Additional information

Based on the absence of data for tetrasodium hydrogen 2-phosphonatobutane-1,2,4-tricarboxylate ("PBTCNa4"), the read-across approach is proposed with 2-phosphonobutane-1,2,4-tricarboxylic acid ("PBTC").

In aqueous media, PBTCNa4 and PBTC dissociate into the corresponding anion (2-phosphonatobutane-tricarboxylate ion) and the sodium ion and hydrogen ion (proton), respectively. Fate, behavior and the ecotoxicological properties of PBTC and its tetrasodium salt are thought to be an effect of the phosphonato-carboxylate ion rather than of the sodium ion or the hydrogen ion (proton), which are normal constituents in environmental systems and have no relevant ecotoxic properties in low concentrations.

Therefore a read-across between PBTCNa4 and PBTC is justified.

In a comprehensive study, the irradiation of [14C]PBTC with simulated sunlight (artificial light source) in a merry-go-round apparatus showed different results. Generally, the formation of a reactive photosystem, probably by complexation of ions [e.g. Fe (III)], was found to be necessary for an absorption of environmentally relevant light (wave length > 290 nm) as well as for a transformation of PBTC to a slightly less polar main photoproduct. Depending on the marginal conditions in the irradiated solutions [pH and ion concentration, e.g. Fe (III)] an equilibrium between PBTC and photoproduct at quite different ratios was to be observed. The most effective transformation at a low level of remaining PBTC in the irradiated solutions was determined for pH 9 in the presence of Fe(III), but still more transformation was observed in pure water stored together with PBTC in a brown glass vessel prior to irradiation. Higher amounts of ions being present in the test solution, e. g. Fe (III), probably decreased the transformation rate or enhanced the back-reaction to PBTC. This influence was already described by Kleinstück [Degradation of phosphonic acids in aqueous solution by light (KSK 4410). BAYER AG, Leverkusen, AC, unpublished report no. AC-6-1252-/m (1990).]. The butane-1,2,4-tricarboxylic acid (BTC) was found to be a final product of PBTC photolysis, but the amounts at the respective sampling periods could not be measured, because a poor chromatographic separation from the main photoproduct. Therefore, no further information about the stability of the main photoproduct could be given. The corresponding irradiation in natural water (Hönniger Weiher) resulted in a PBTC half-life of about 30 minutes and a PBTC steady state concentration of less than 10% of initial after 420 minutes. This proved that photolysis can contribute to the overall degradation of PBTC in natural waters. The estimates of "environmental photolysis half-lives" based on two different arithmetic models (GC-SOLAR and Frank & Klöpffer) by means of the resulting quantum yields and the light absorption data in the environmentally relevant range of wavelengths were well comparable when considering identical marginal conditions. The results of modelling based on the irradiation of PBTC in buffer pH 9 in presence of FeCl3 indicated that the mean photolysis half-life should range from 2-3 days in summer to 15-65 days in winter. The results of modelling based on the irradiation of PBTC in pure water, stored in brown glass prior to irradiation indicated that the mean photolysis half-life should range from 0.2-0.3 days in summer to 1-10 days in winter. The before-mentioned assessments did not consider any indirect photodegradation mechanisms which could increase the degradation in natural waters.

For a computer-based assessment of the so-called environmental photolysis half-life in water the quantum yield of direct photoreaction of [3,4-14C]PBTC, should be determined according to the ECETOC method in polychromatic light.

The irradiation of [14C]PBTC with simulated sunlight (filtered Xenon light) in a merry-go-round apparatus showed different results. Generally, the formation of a reactive photosystem, probably by complexation of ions [e.g. Fe(III)], was found to be necessary for an absorption of environmentally relevant light (yield > 290 nm) as well as for a transformation of PBTC to a slightly less polar main photoproduct.

Depending on the marginal conditions of irradiation [pH and ion concentration, e.g. Fe(III)] an equilibrium between PBTC and main photoproduct at quite different ratios was observed. An effective transformation at a low level of remaining PBTC in the irradiated solutions was determined for pH 9 in the presence of Fe(III). The quantum yield was calculated to be 0.0022. More effective was the transformation in pure water being stored together with PBTC in a brown glass vessel prior to irradiation. In that case, only traces of e.g. Fe or Mn ions can have been dissolved, but the quantum yield was calculated to be 1.84. Higher amounts of ions being present in the test solution, e.g. Fe(III), probably decreased the transformation rate or enhanced a back-reaction to PBTC.

The butane-1,2,4-tricarboxylic acid (BTC) was found to be a final product of photolysis of PBTC. The quantities of BTC at the respective sampling periods could not be determined, because of a bad chromatographic separation from the main photoproduct. Therefore, no further information about stability of the main photoproduct could be given.

The corresponding irradiation of [14C]PBTC in a natural water (Hönniger Weiher, NRW) resulted in a PBTC half-life of about 30 minutes and a PBTC steady state concentration of less than 10% of the initial concentration of 1 mg/L. This proved that photolysis can contribute to the overall elimination of PBTC in natural waters. The measured fast photo-transformation was unexpected, because not any absorption of light above 230 nm was measurable. Therefore, not any calculation of a quantum yield according to the above-mentioned method was possible. Nevertheless, the acting quantum yield must have been extremely high (yield >> 1), because that fast photo-transformation was measured.

The estimates of "environmental photolysis half-lives" based on two different arithmetic models (GC-SOLAR and Frank & Klöpffer) by means of the resulting quantum yields and the light absorption data in the environmentally relevant range of wavelengths were well comparable when considering identical marginal conditions. The results of modelling based on the irradiation of PBTC in buffer pH 9 and in presence of FeCl3 indicated that the mean photolysis half-life should range from 2-3 days in summer to 15-65 days in winter. The results of modelling based on the irradiation of PBTC in pure water and stored in brown glass prior to irradiation indicated that the mean photolysis half-lives should range from 0.2-0.3 days in summer to 1-10 days in winter.